Dietary treatment postpartum in women with obesity reduces weight and prevents weight gain: a randomised controlled trial

In women of reproductive age, obesity is becoming more prevalent worldwide [1]. Obesity is an important risk factor for multiple co-morbidities [2], such as cardiovascular disease (CVD) [3], the leading cause of death and one of the most frequent causes of disability-adjusted life years lost in women [4]. Obesity in women may partly be explained by the association between parity and weight gain during midlife [5]. Increased parity is also correlated with increased intra-abdominal adipose tissue [6] and the metabolic syndrome [7].

Pre-pregnancy obesity increases the risk of adverse outcomes in both baby and mother during pregnancy and birth [8, 9], in addition to an elevated risk of maternal CVD and diabetes type 2 in the future [10]. Women with higher pre-pregnancy BMIs tend to retain or gain more weight postpartum and increase central fat mass more than leaner women [11‐13]. Hence, women with obesity may enter a vicious cycle during reproduction, risking an even more pronounced overweight with associated risk factors for each gestation. Increasing weight gain between pregnancies has been shown to elevate the risk of infant mortality [14] and maternal complications [15] in the subsequent pregnancy. The first six months postpartum seem crucial in order to minimise postpartum weight retention (PPWR), as failure to lose excess pregnancy weight by this time point is an indicator of increased BMI later in life [16]. In addition, many women with overweight and obesity are motivated both to lose pregnancy weight in the postpartum period [17] as well as to serve as role models for a healthier lifestyle [18]. Energy requirements increase during breastfeeding and may therefore facilitate weight loss [19‐22]. A weight loss of 0.5 kg per week does not seem to affect breastmilk production or quality [23, 24].

Interventions including diet alone, or in combination with exercise, have shown to reduce weight in postpartum women, but studies often include none or few women with obesity and rarely from obesity classes with a BMI > 35 kg/m² [25, 26]. Bertz et al. have demonstrated that a 12-week structured dietary behaviour modification treatment using the LEVA (Lifestyle for Effective Weight loss during Lactation) method resulted in a weight loss of 9 kg (10%) 12 months after intervention in postpartum women with a pre-pregnancy BMI of 25–35 kg/m² [27]. The dietary treatment led to favourable effects on fasting insulin and HDL cholesterol (HDL-C) [28]. The method for weight loss has also been shown to be effective within the primary health care setting [29]. These results imply a great potential for the use of this method also in women with BMI > 35 kg/m². However, to what extent the LEVA method is effective in producing clinically relevant weight loss and improving the cardiometabolic risk profile in these women needs to be determined.

The aim of this study was to investigate the effect of a dietary treatment using the LEVA method in postpartum women recruited from the specialist outpatient clinic Adipol in Oslo, Norway, following pregnant women with obesity. The primary outcomes were changes in body weight and markers of lipid and glucose metabolism after 3 and 12 months. Secondary outcomes were changes in PPWR, body composition, blood pressure, waist and hip circumference, high-sensitivity C-reactive protein (hs-CRP) and treatment-related changes in dietary intake and physical activity.

In the present randomised controlled trial, we aimed to test the effects of a diet intervention using the LEVA method on weight loss and cardiometabolic risk factors in postpartum women with higher BMIs than in previous studies testing the same intervention. The women had an average BMI at baseline of 40.0 kg/m², compared to the previous studies, in which mean baseline BMIs were 30.1 kg/m² [27] and 31.7 kg/m² [29]. The diet treatment was effective in reducing weight, waist circumference and PPWR after both 3 and 12 months. After 3 months, the dietary treatment reduced fat mass, and after 12 months, fat free mass and fasting glucose concentration declined as an effect of dietary treatment. Previous published studies on the LEVA method showed a weight loss at 12 months that was more than twice as large (9 kg, 10%) [27, 29] than in the current study (4 kg, 4%). However, the unadjusted between-group difference at 12 months in the present study was 9 kg (7 percentage points), which is similar to the study by Bertz et al. of 7 kg (8 percentage points) [27] and in the effectiveness trial by Huseinovic et al. of 6 kg (7 percentage points) [29].

The goal of the programme is to lose 6 kg in 12 weeks, which only one participant in the diet group achieved. However, the mean GWG in the present study was half the size as in the effectiveness study (8 kg versus 17 kg) [29]. The mean PPWR was only 1 kg at baseline, which leaves less pregnancy weight to lose. Still, we observed significant between-group differences in weight change from baseline to 3 and 12 months and PPWR in the present study. These differences are partly caused by a weight gain in our control group (within-group changes not statistically tested), as opposed to the weight loss seen in the control groups in the previous studies [27, 29]. The increased PPWR in our control group is therefore not only a result of an actual retention of pregnancy weight, but also a weight gain in the postpartum period. Our finding is in line with observational studies, as women in higher BMI-categories tend to gain or retain weight during the postpartum period [11, 12]. The reasons for and mechanisms behind these observations are less discussed in the literature, but the prevalence of antenatal depression symptoms and eating disorders is higher in women with pre-pregnancy obesity than in women with a BMI in the normal range [38]. Weight gain from pre-pregnancy to 6 months postpartum is associated with risk of depression and anxiety, however, causal direction may go both ways [39]. Obesity in adults is associated with binge-eating disorder [40], for which emotional eating, i.e., an urge to eat in response to emotional rather than physical cues, is a risk factor [41]. None of the mentioned conditions or symptoms were examined in the present study, but considering the potential for emotional stress in the postpartum period, a weight gain in this population in response is not unexpected. Most et al. found that weight gain in postpartum women with obesity was a result of increased energy intake rather than decreased energy expenditure or differences in breastfeeding [42]; They further reported that concentrations of appetite-regulating hormones like leptin and cholecystokinin were higher in the women gaining weight compared to the women losing weight postpartum. In the current trial, energy intake was not statistically significantly reduced by treatment, despite the significant effect of weight. This may be explained by the reduced reported energy intake in controls after 12 months and large variation in the data. Additionally, the FFQ’s ability to detect changes in energy intake is not validated. The findings in the present study support observational studies showing that excess pregnancy weight is not necessarily lost naturally in women in higher BMI classes; rather, the opposite may occur [11, 12]. Our data suggest that weight gain postpartum may be prevented and even turned into weight loss when an effective dietary intervention is provided.

Two meta-analyses reviewing trials with lifestyle interventions for weight loss postpartum reported a pooled between-group difference of only around 2.5 kg [25, 26]. However, the use of self-monitoring and the combination of diet and physical activity, rather than physical activity alone, increased weight loss [25]. We used the LEVA method, which is designed to build self-efficacy through self-monitoring with regular feedback on weight loss goals. Dietary change is the key component of the method, with limited focus on physical activity, as a more extensive exercise intervention failed to preserve more fat-free mass compared to the control group [27]. A review on lifestyle interventions from 2018 identified individualised support as a criterion for successful weight loss in postpartum women [43]. The LEVA method provides personal and concrete guidance by a dietitian, who together with the participant sets goals and discusses challenges and solutions [27, 44]. In the current study, the intervention using the LEVA method succeeded in reducing weight and PPWR after 12 months in a group of women with pre-pregnancy obesity and a high risk of long-term excess PPWR. However, group-specific tailoring of the method may be beneficial. For example, a lower and more realistic weight loss target may be favourable, in order to maintain self-efficacy.

The reduction in weight, waist circumference, PPWR and fasting glucose caused by the diet intervention may be clinically relevant despite the lack of statistically significant effects on the other markers of lipid and glucose metabolism. Waist circumference and abdominal obesity are associated with early atherosclerosis in young adults [45], as well as acute myocardial infarction [46]. An increase in BMI of > 1 BMI unit from pre-pregnancy to 18 months postpartum is associated with a higher risk of hypertension [47]. Failure to lose weight after pregnancy is a predictor of higher BMI 10 years later [16], which may increase risk of co-morbidities like CVD [3]. Adverse maternal outcomes increase in the second pregnancy of women with an increasing BMI between the first and second pregnancy [48], among them preeclampsia, which is known to increase the risk of maternal CVD later in life [49]. The dietary treatment reduced fasting blood glucose at 12 months, and when excluding the two women in the control group who lost a substantial amount of weight participating in external diet programmes, the diet intervention led in addition to lower insulin levels. Both findings indicate an improvement in glucose metabolism, and potentially in insulin sensitivity, which is favourable for this group of women with increased risk of type 2 diabetes [10]. Although not statistically significant, the size of the positive effect of the dietary treatment on HDL-C at 12 months was similar to that observed at 1 year in the LEVA study [28]. The changes seen in the diet group were in a more favourable direction for most of the markers of lipid and glucose metabolism compared to the control group, although statistically non-significant. The lack of statistical significance may be due to too small effect sizes, or too large variation of these, relative to the sample size.

Our study has limitations. We aimed to include 40 subjects but terminated before we reached this number because of stagnation in the number of eligible women at the clinic. In addition, the 12-month lost-to-follow-up rate was higher (45%) than expected (25%). With more study subjects, we would consider to adjust for breastfeeding status at each visit, as breastfeeding may influence body weight [50] and markers of metabolism of glucose [51] and lipids [52]. We performed a number of statistical tests; however, we decided not to correct for multiple testing as many of the outcomes are related and are the results of similar mechanisms. All blood samples for analysis of fasting glucose were centrifuged between 30 and 60 min after drawing. This procedure was according to the recommended routines of the accredited laboratory that analysed the blood samples, but the samples may still have been affected by glycolysis. However, we find no risk for systematic bias between treatment groups concerning time point for centrifugation. Physical activity level was examined using a non-validated questionnaire, adjusted to capture the activity in a postpartum setting. A short version of a similar questionnaire consisting of only two questions has been validated and considered sufficiently reliable and valid at group level [53], but has not been validated in postpartum women or to measure changes in physical activity levels.

The strengths of our study are the randomised, controlled design, the use of an evidence-based dietary intervention and the inclusion of subjects that are less studied. Our findings are applicable to postpartum women with obesity and a personal motivation to lose weight who have the benefit of a comparatively long maternity leave and who have a higher education level than average Norwegian women [54].

The LEVA method has been shown not only to produce clinically relevant postpartum weight loss at 2 years but also to increase quality of life and be a cost-effective treatment [55]. The postpartum period is a window of opportunity for women with overweight and obesity to obtain a healthier lifestyle and lose weight. This may also contribute to improved health behaviours in the whole family, in addition to potential favourable effects in subsequent pregnancies. With the potential to reduce the intergenerational cycle of obesity and associated co-morbidities, there is an urgent need for evidence-based recommendations, policies and specific actions, as well as trained health care professionals, in this area of maternal health [43].

We have shown that the dietary treatment group decreased weight, waist circumference and postpartum weight retention compared to the control group in postpartum women with obesity. The dietary treatment may prevent postpartum women with obesity from retaining or gaining weight and may even reduce weight before subsequent pregnancies and later in life.